Source of Lead in Central American and Caribbean Mineralization, II

Home , Chortis Block

Earth and Planeta O, Science Letters, 56 ( 1981 ) 199- 209 199 Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

[2]

Source of lead in Central American and Caribbean mineralization, II. Lead isotope provinces George L. Cumming Institute of Earth and Planetary Physics and Department of Physics, Universi O' of Alberta, Edmonton, Alta. ToG 2J1 (Canada) Stephen E. Kesler Department of Geological Sciences, UniuersiO, of Michigan, A n n Arbot; MI 48109 ( U. S. A.)

and Dragan Krstic Department of Pl~vsics, Universi(v of AIherta, Edmonton, AIta. T6G 2J1 (Canada)

Received March 13, 1981 Revised version received July 7, 1981

In an earlier study of Mesozoic and Cenozoic mineralization in Central America and the Caribbean region, we found that lead isotopic compositions of deposits in northern Central America, which is underlain by a pre-Mesozoic craton, ranged to higher 2°6pb/Z°aPb and 2°7pb/Z°4pb compositions than did deposits from elsewhere in the region, where the basement is Mesozoic oceanic material. Using 16 analyses for 12 new deposits, as well as new analyses for 11 of the samples studied previously, we have found that lead isotopic compositions correlate closely with crustal type but show little or no correlation with depth to the M-discontinuity. The deposits are divisible into three main groups including (in order of increasing 2°7pb/2°4pb and 2°8pb/Z°4pb ratio): (1) deposits in southern Central America and all deposits in the Greater Antilles except Cuba; (2) all deposits in northern Central America; and (3) the Cuban deposits. Southern Central American and Caribbean lead is higher in 2°vpb/Z°4pb and 2°spb/2°4pb than most mid-ocean ridge basalts but could have been derived directly or indirectly from undepleted mantle. Northern Central America can be divided into the Maya block, which belongs to the Americas plate, and the Chortis block, which belongs to the Caribbean plate. Maya block deposits fall along a linear array whereas those of the Chortis block (except the Monte Cristo deposit) form a cluster. These results suggest that the Maya block is underlain by crust or mantle with a large range of U/Pb and Th/U ratios, whereas the Chortis block basement is more homogeneous. Two-stage model calculations indicate an age of about 2280-+- 310 m.y. for the Maya block basement, although no such rocks are known in the region. Comparison of the Chortis block data to our recently published lead isotopic analyses of Mexican deposits shows considerable similarities suggesting that the Chortis block could have been derived from Mexico.

I. Introduction possible to recognize smaller groupings of deposits with close similarities in both lead isotopic com- In an earlier study of Mesozoic and Cenozoic position and geologic environment. In particular, mineralization in Central America and the Carib- we found that the lead isotopic compositions of bean region [1] we noted that although most de- deposits from northern Central America, which is posits exhibited generally similar lead isotope underlain by a pre-Mesozoic craton (Fig. 1), ranged compositions approximately linearly related, it was to higher ratios than did deposits from elsewhere

0012-821X/81/0000-0000/$02.75 ,(: 1981 Elsevier Scientific Publishing Company 200

0 200 400 km ie

Pre Mesozac Continental Crust

m Oceanic Crust

- - Depth to M-Olscontlnulty (km)

Fig. I, Map of Central America and the Caribbean region showing distribution of prr-Mesozoic continental crust as well as oceanic crust, largely of Mesozoic age. Depth to M-discontinuity is also shown. after Case [20]. Locations of analyzed deposits are indicated by numbers that correspond to numbers in Tables I and 2.

in the region, where the basement is Mesozoic study. The analysis on a sample from the Pueblo oceanic material. In a recent summary of such Viejo gold deposit in the Dominican Republic, work, Doe and Zartman [2] have shown that there reported in our previous paper, has been su- is indeed a close relation between the type of perseded by an extensive study of that deposit, the basement rocks underlying an area and the iso- results of which will be reported elsewhere. The topic characteristics of lead in ore deposits of the remaining four samples were not remeasured since area. Stacey et al. [3], Zartman [4] and Stacey and duplicate analyses agreed within our present mea- Zartman (51 have used the lead isotopic composi- suring error limits. The new results presented here tion of Mesozoic and Cenozoic mineral deposits to indicate a clear correlation between the lead iso- characterize and date Precambrian and younger tropic composition of a mineral deposit and its rocks concealed beneath the deposits. In order to crustal environment. evaluate more completely the relation between crustal environment and the lead isotopic composition of the widespread Mesozoic and Cenozoic 2. Geologic setting of mineralization in the Carib- ore deposits of the Caribbean region, we have bean region obtained 16 new lead isotopic analyses from twelve 2.1. Introduction previously unstudied deposits and district (Table 1 and 2; Fig. 1) and have reanalysed eleven of the The Caribbean region (Fig. 1) includes four is- fifteen samples that were included in our original land arc systems: Central America, the Greater 201

Antilles, the Lesser Antilles, and the islands north cept for a possible metamorphic core beneath of South America. Although the distribution and western Cuba. types of mineral deposits differ from arc to arc [6] most of this variation is seen in Central America 2.3. Mineral deposits and the Greater Antilles to which we have con- fined this study. The Greater Antilles developed All of the Central American mineral deposits by interaction between the Americas and Carib- included in this study are of Cenozoic or probably bean plates in Mesozoic and early Cenozoic time, Cenozoic age [6]. In northern Central America, whereas Central America interacted with the Cocos deposits in the Maya block and in the Polochic- plate, at least in Cenozoic time [7]. Motagua fault zone, contain lead-zinc, are hosted by Cretaceous and Paleozoic limestones, and show 2.2. Basement rock similarities to the Mississippi Valley-type deposits of North America [21]. Deposits south of the fault Northern Central America is underlain by an zone in the Chortis block of northern Central extensive pre-Mesozoic craton that is divided into America include: (1) silver-lead-zinc limestone re- two parts by the Polochic-Motagua fault zone that placement deposits (Mochito, Las Animas, La is considered to be the Americas-Caribbean plate Vueltosa, Las Minas) in Cretaceous limestone; and boundary [8]. The presence of ophiolites in this (2) silver-gold-lead-zinc veins in Cenozoic volcanic zone [9] supports the suggestion from paleo- rocks (Monte Cristo, Bonanza, E1 Rosario). magnetic data [10,11] that northern Central As noted by Cumming and Kesler [1], lead America actually consists of two pre-Mesozoic mineralization and trace lead are less abundant in crustal fragments that are separated approximately southern Central America and the Greater Antil- by the Polochic-Motagua fault zone. Dengo [12] les. In southern Central America, the only galena- recognized differences between the pre-Mesozoic bearing samples available to us are from late vein- rocks on opposite sides of this zone and called the lets in the large Cerro Colorado porphyry copper northern part the Maya block and the southern deposit and the Boston gold-silver vein deposit in part the Chortis block (Fig. 1). Costa Rica. Both deposits are late Cenozoic in age No such metamorphic basement complex is [22,23]. In the Greater Antilles, galena is found in present in southern Central America or the Greater volcanogenic massive sulphide deposits of Jurassic Antilles. Dengo [13] and del Guidice and Recchi age (Cuba), early Cretaceous age (Haiti) and early [14] report ophiolite complexes along western Costa Cenozoic age (Jamaica) [6], as well as probable Rica and Panama (Fig. 1) and similar rocks are early Cenozoic vein deposits in Puerto Rico [24]. present from eastern Cuba to Puerto Rico in the An important aspect of the samples included in Greater Antilles [15-18]. Small exposures of phyl- this study is the fact that they range in age from litic rocks similar to those in nuclear Central Jurassic to Pliocene and are found in host rocks America are found in westernmost Cuba [19]. ranging from probably coeval intrusive, volcanic, The approximate extent of the metamorphic or sedimentary units, to genetically unrelated sedi- craton in northern Central America is shown by mentary units ranging from limestone to shale and the 30-km depth contour to the M-discontinuity sandstone. This fact, along with the observation [20] as shown in Fig. 1. Throughout southern that ore leads commonly exhibit a wider composi- Central America, the crustal thickness is consider- tional range than related, igneous rock leads [4, ably less than 30 km except in northern Costa figs. 1 and 2], suggests that our results may cover Rica (Fig. 1). No metamorphic core has been iden- essentially the entire range of isotopic composi- tified in this area even though the crust is thicker tions to be observed in the Mesozoic and Cenozoic there. Crustal thickening may be related to the igneous rock and ore leads in this region. fault zone which forms the southern boundary of the Chortis block, according to Dengo [12]. Thin crust prevails throughout the Greater Antilles ex- 202

3. Results and discussion galena, but some pyrite and sphalerite were also analysed and these samples are indicated in the 3.1. Experimental techniques table. The dhta of Table 1 can be divided into three The data of our previous paper [1] were essen- main groups (Fig. 2) on the basis of their tially of a reconnaissance nature and experimental 207 pb/204 Pb and 208 pb/204 Pb ratios. These are: (1) procedures have improved considerably since that the deposits in southern Central America and all time, hence we have reanalysed most of the origi- deposits in the Greater Antilles except Cuba; (2) nal samples in order that they be directly com- all deposits in northern Central America; and (3) parable with our new analyses. These new results the Cuban deposits. We note further that although were obtained over a three-year period and several the data from the Maya and Chortis basement samples were analysed repeatedly in order to check blocks overlap (group 2 above), the Maya block the long term reproducibility of our measure- data form an extended linear array, similar in ments. 2°6pb/2°4pb ratios but enriched in 2°Tpb/z°4pb as The measurements were carried out on a Micro- compared with group 1 data, whereas the data mass MM30 mass spectrometer using a rhenium from the Chortis block form a much shorter and filament and the silica gel/phosphoric acid load- less radiogenic array. In fact, if we exclude sam- ing technique. The 1- to 2-lxg sample was brought ples from the Monte Cristo mine in E1 Salvador to an operating temperature of 1400°C as mea- and the sample from the Montenegro deposit in sured by a disappearing filament optical pyrome- Guatemala, the remaining Chortis block data form ter and data were taken after an approximately a cluster which is scarcely larger than our experi- one-half hour waiting period. Under these experimental error. mental conditions the fractionation correction was Adequate explanation of the position of the 0.5%0 per mass unit difference. The results of 40 Monte Cristo analyses is important to a full char- replicate measurements of NBS SRM 981 indicate acterization of the isotopic characteristics of the a measuring precision of about 0.4%o per mass Chortis block. Unfortunately however, not enough unit difference (2o), and a correlation coefficient is known about the age and petrology of the of about 0.75 between pairs of ratios. volcanic rocks in the Monte Cristo area to de- Although the experimental procedures varied termine its relation to other deposits in the Chortis during the course of this study, we have not re- block. In fact, the Monte Cristo deposit and its jected measurements except where older measure- host volcanic sequence are most similar to the ments have been superseded by more recent val- numerous lead-poor, precious metal deposits of ues. The data of Table 1 show only the average southern Central America, which are represented values. The numbers of replicate analyses are indi- in this study only by the Boston mine (sample 7, cated in Table 2. We feel that the above errors are Fig. 2). Thus, it is likely that further data are in fact realistic, since when various groupings of needed on these deposits in order to account ade- data are used to calculate least squares straight quately for the Monte Cristo analyses. In contrast, lines, the mean squared weighted deviates (MSWD) all other deposits in the Chortis block for which are always close to one even though errors have data are available appear to be part of a single been reduced by a factor corresponding to the geological group which formed in mid-Cenozoic number of replicates when more than one mea- time, and which has a very restricted range of surement was averaged for a particular sample. isotopic compositions. Comparison of the isotopic data for each de- 3.2. Lead isotopic results posit with the depth to the M-discontinuity at that locality indicates that there is no simple correla- Our analytical data are shown in Table 1, di- tion between isotopic composition and crustal vided into groups according to their geologic and thickness. Note for instance that the Boston mine geographic environment. Most samples were in Costa Rica has a moderately high 2°6pb/Z°4pb 203

TABLE 1 Lead isotopic composition of sulphide minerals from Mesozoic and Cenozoic mineral deposits in the Greater Antilles and Central America. The samples are divided into groups on the basis of their geographic and geologic settings and are listed where possible, in order of decreasing geologic age (sample locations are shown in Fig. 1) Sample No. 2o6 Pb/204 Pb 207pb/2O4 Pb 20s pb/2O4 Pb

Greater Antilles

1 CBA-SL 18.693 15.685 38.828 2 CBA-CAST 18.714 15.712 38.892 3 H-K-1 19.048 15.621 38.849 4 H-K-2 19.062 15.630 38.859 5 PR-72-GA 18.877 15.606 38.646 6 J-72-HM 18.772 15.593 38.476

Southern Central Ameriea

7 CR-AB 18.901 15.599 38.710 8 78-722 19.054 15.627 38.804 9 P-CC 19.096 15.605 38.812

Northern Central America

Maya block l0 VL-SJ 18.939 15.668 38.867 I I RB-GA 19.038 15.693 39.085 12 A-27 18.816 15.653 38.774 13 SD-3 18.831 15.668 38.790 14 A-143 18.870 15.670 38.855 15 A-195 18.873 15.678 38.871 16 TZ-3 18.737 15.642 38.685 17 A-490 18.751 15.649 38.676 18 A-467 18.821 15.652 38.693 19 A-486 18.870 15.659 38.789

Chortis block 20 PNCO-4-8 18.771 15.637 38.661 21 MTNG-MTO 18.801 15.654 38.714 22 H-R 18.765 15.654 38.716 23 H-LA 18.766 15.639 38.678 24 H-LV 18.776 15.634 38.650 25 ML-1225-SJ 18.777 15.649 38.703 26 N-VS- I 18.762 15.652 38.729 27 N-VS-2 18.748 15.637 38.663 28 BLAG 18.760 15.638 38.654 20 LL-I-ES 18.663 15.617 38.481 30 ES-2-MC 18.655 15.621 38.535

ratio with a low 2°7pb/2°4pb ratio characteristic of probably made up in part of Paleozoic rocks. Thus the oceanic crust group (Fig. 2), and yet is under- lead isotopic composition may be more closely lain by crust that is over 40 km thick (Fig. 1). On correlated with crustal type than with crustal the other hand the Monte Cristo deposit comes thickness alone. from an area of relatively thin crust and yet has a Present models for the plate tectonic evolution higher 2°7pb/Z°4pb ratio coupled with the lowest of the Central American-Caribbean region [7,25] 2°6pb/2°4pb ratio of any of the samples in this indicate that there are no suitable DSDP locations group. The crust below the Monte Cristo deposit is which could represent ocean floor material that 204

TABLE 2 Notes on the analyses of Table 1 (for samples analysed in duplicate, mean values are given) Sample Number of Mineral(s) Country District Deposit Type No. duplicates Greater A ntilles

I 4 P, G Cuba Pinar dcl Rio Santa Lucia massive sulphide 2 2 P, G Cuba Pinar del Rio Castillanos massive sulphide 3 4 P, G, S Haiti Camp Coq massivc sulphide 4 3 P, G, S Haiti Camp Coq massive sulphide 5* 1 G Puerto Rico 6* 1 G Jamaica Hope Hope massive sulphide Southern Central America

7 4 G, S Costa Rica Abangarcs Boston gold-silver vein 8 5 S Panama Cerro Colorado Cerro Colorado porphyry_' copper deposit DDH-78,722 foot depth 9 5 P, G, S Panama Cerro Colorado Cerro Colorado peripheral vein at surface

Northern Central America

Mt(va block 10" O Guatemala San Miguel Villa Linda limestone replacement 11" G Guatemala San Miguel Rosario limestone replacement 12" G Guatemala Chiantla Santo Domingo limestone replacement 13" G Guatemala Chiantla Santo Domingo limestone replacement 14" G Guatemala Chiantla Torlon vein 15" G Guatemala Chiantla Torlon vein 16"* G Guatemala Chiantla Tziminas limestone rcplaccment 17 ** G Guatemala Coban Suquinay limestone replacement 18 ** G Guatemala Coban San Joaquin limestone replacement 19" G Guatemala Coban Caquipec limestone replacement

Chortis block 20 3 G, S Guatemala Las Minas Pefiasco limestone replacement 21 G Guatemala Las Minas Montenegro limestone replacement 22 G Honduras San Juancito Las Animas limestone replacement 23 G Honduras San Juancito Las Animas limestone replacement 24 G Honduras Mochito La Vueltosa limestone replacement 25 ** G Honduras Mochito San Jorge limestone replacement 26 G Nicaragua Bonanza Neptuna vein 27 G Nicaragua Bonanza Neptuna vein 28 2 G Nicaragua Bonanza Blag gold-silvcr vein 29 * I G El Salvador Fonseca Monte Cristo silvcr-gold vein 30 4 G El Salvador Fonseca Monte Cristo silver-gold vein G=galena, S=-sphalerite, P=pyrite. * Reanalysis of samples previously reported in Cumming and Kesler [1]. ** Previously published analysis [1].

might have undergone subduction and partial such possible lead sources as the mantle, the lower melting in mid-Cenozoic time to produce the ob- or upper crust, and ocean floor basalt and sedi- served isotopic compositions. Thus, it is not possi- ment. It is clear, however, that the majority of data ble to use direct mixing models such as that of on rock leads from mid-ocean ridge basalts are Church [26] to evaluate the relative importance of considerably lower in 2°Tpb/2°apb and 2°spb/ 205

NORTHERN CENTRAL AMERICA Central America, however, and a source with higher x MAYA BLOCK • CHORTIS BLOCX • MONT~ CPdSTO U/Pb and Th/U ratios must be sought. Although SOUTHERN CENTRAL AMERICA A AMERICA this source could have been either continental 3g.z T GREATER ANTILLES (9 CUBAN MASSIVE SULFIDE DEPOSITS J material or unusual subcratonic mantle, the small • OTHER DEPOSITS ~ / ~ t, size of the Central American craton and its ap- parently wide range of Mesozoic and Cenozoic {[ o -'" movement [30] make a crustal source seem more likely. In accordance with the generalizations of

16x 17 a Zartman [4] regarding the characteristics of lead isotopic analyses from the western United States,

i119 . 2 Slime £rrors it is concluded that the lead forming the linear array in the Maya block was derived from a crustal 188 or mantle source containing a substantial range of 15 7 U/Pb and Th/U ratios. In contrast, the Chortis block lead could have been derived from a more an IBx i al~° homogenized source with a limited range of U/Pb a ~ 9~ and Th/U ratios. Alternatively material could have been homogenized at the time of ore formation, but in either case the source of lead is different for 1 2 Sillma Errors I ~ the Chortis and Maya blocks. 15.5 18.a 18 7 18 S 189 lg0 1~ 1 The Cuban lead exhibits extreme 2°7pb enrich- ~Pb / ~P b ment, possibly reflecting a crustal source for the Fig. 2. Isotopic compositions of lead in Central American and Jurassic sediments that host them [19, p. 39]. It Caribbean mineral deposits. Growth curve shown is that of should be noted that although the two Cuban Stacey and Kramers [42]. Lines 1 and 2 represent best-fit lines deposits appear to be the same within experimen- for two of the three major divisions of the data (1 =southern tal error, the duplicate measurements are always Central America and the Caribbean region (except Cuba); 2=northern Central America). Area 3 represents the Cuban clearly distiiict from one another; that is, the ratios data. for sample 2 were always higher than the ratios for sample 1. We are thus confident that the two deposits differ in composition from each other and thus the isotope ratios reflect real differences in 2°4pb than the data presented here (see Tatsumoto the source of these two leads. This, coupled with [27,fig. 9]). Only our group 1 data from southern the extremely high values of 2°Tpb/2Capb, suggests Central America and the Greater Antilles overlap an enriched and variable crustal source which must the range of compositions for any oceanic rocks have been differentiated from more normal crust and these are the rather atypical data from DSDP at an early age in order to produce the observed Leg 37 [28,29], and from Gough and Reunion spread in the 2°Tpb/2°4pb ratios. islands (see Tatsumoto [27, fig. 11]). By analogy with models adopted for the oceanic island basalts 3.3. Straight line fitting and model age calculations [27], relatively undepleted mantle is the only com- positionally acceptable material that could have We have calculated least squares straight lines supplied lead to the Caribbean and southern [31] for the various data groups (Fig. 2) as indi- Central American deposits. Presumably this could cated above, and used these slopes to determine have taken place by subduction in the Caribbean possible ages of source rocks assuming a two-stage region, of mid-ocean ridge basalts derived from model. Since ages of formation of the various ore such mantle. bodies range from Jurassic to Pliocene, this is Even lead from these oceanic sources is dis- clearly an oversimplification, but the results will tinctly different from the data from northern not be critically dependent on this approximation. 206

For calculation purposes we have used a value for 4. Relation of Caribbean and Mexican lead iso- t 2 of 30 m.y. (mid-Cenozoic) in the two-stage topic data and source of the lead anomalous lead equation: exp( X't, ) -- exp( X't 2 ) Inasmuch as the Cenozoic tectonic processes R: that generated the Central American mineraliza- a [exp(Xt, ) - exp(2tt 2 )] tion extended northward into Mexico and, in fact, where R is the slope of the line and the decay were most extensive there, it is reasonable to com- constants and uranium isotope ratios are as recom- pare the new Central American analyses to our mended by Steiger and J~iger [32]. recently published data from Mexico [37]. As can The eleven ratios for samples from the Chortis be seen in Fig. 3, there is considerable overlap block yield a line of slope R = 0.186-+ 0.030 and between data from Mexico and northern Central MSWD = 1.33. From the Maya block, ten samples America, although the Central American Maya fit a line of slope R - 0.146 + 0.029 and MSWD = block analyses extend to higher ratios and the 0.98. These slopes yield ages of 2680-+-250 m.y. Mexican analyses extend to lower ratios. An im- and 2280-+ 310 m.y. (lo) respectively, with rela- portant characteristic of the isotopic data from tively large errors due to the short lines. The most of the larger deposits in Mexico was their combined data sets from the Maya and Chortis uniformity of isotopic composition. Analyses from blocks form a line of slope R = 0.195 + 0.015 with important mineral districts such as Taxco, Fres- MSWD = 1.31. Although the error for this line has nillo, and Parral (Fig. 3) formed clusters similar to decreased considerably, the age (t~ : 2780-+ 120 those observed for the Chortis block analyses. As m.y.) and slope are even greater than those of the can be seen in Fig. 3, the composition of these individual groups. The ages indicated for the source Mexican clusters is very similar to that of the of the Central American ores from these calcula- Chortis block. Furthermore, the stippled area sur- tions are clearly much greater than the oldest rounding these clusters includes almost all of the radiometric dates in the area. Gomberg et al. [33] mineral deposits in northern Mexico. It is likely, obtained ages of 1000 m.y. on zircons from the therefore, that the same process or crustal in- metamorphic core of the Maya block. Rocks of fluence prevailed during formation of the lead in great age are also lacking in southern Mexico [34] most of the Mexican 0eposits and in those of the and northwestern South America [35], but older Chortis block. Those deposits with analyses falling rocks are present in northeastern South America outside the stippled zone come largely from south- [36]. ern or northernmost Mexico (Fig. 4) where it is The data from southern Central America and possible that uranium-depleted, lower crustal the Greater Antilles also form a colinear group, if material was a source of the lead. Limited data we exclude the late vein sample (P-CC) from Cerro available to us (analyses l, 2, 3, 4, 5, l0 in fig. 3 of Colorado. We obtain a slope of R = 0.143-+ 0.016 Cumming et al. [37]), as well as the conclusions of with MSWD = 1.49. This yields an apparent age Ewing [38] regarding lead isotopic data from New from the source rock for these ores of 2250 - 180 Mexico, suggest: that the data from northernmost m.y. (la), very similar to the Maya block data Mexico fall in a linear array, indicating a t~ of discussed above. Including sample P-CC in the about 1.5 b.y. for a t 2 of mid-Cenozoic age, and calculation reduces the slope to R = 0.032 -+ 0.017 consistent with the known age of the basement in but with MSWD = 19.4, thus indicating that this that area. We are thus led to conclude that if sample does not fit the pattern of the remaining Mexican lead ores appear to reflect the age of analyses of this group. This discrepancy is not due underlying basement rocks, then the Central to experimental error. All measurements on sam- American lead data could also indicate the ages of ple P-CC had higher 2°6pb/2°4pb and lower the source rocks, although these rocks have not yet 2°7pb/2°4Pb ratios than those measured for sample been identified in the field. It is interesting to 78-722 from within the deposit proper. speculate that if the lead isotope ratio patterns for these ores yield information about ages of underly- 207

CB Chortis Block 39.0 Mexico ~~ F Fresnillo P Porrol T Taxco ~.~.

388

38.6

.--" ~L-~TM Ill'"' 38.4 1B.6 18.7 18.8 189 190 191

15.7

...... ----~CENTRAL AMERICA-GREATER a. 15.6 eo ,,, "'1111111111lilJlll 15.5 , , ...... ,,,,,,, ,,, J,J JlllJtlllllllJIlllllJlo ,, i i i 186 18.7 188 18.9 190 191 206 Pb/204 Pb

Fig. 3. Compositional fields for lead isotopic data from Mexico, Central America and the Caribbean region compared to lead in typical oceanic basalts [27]. Specific clusters in the Mexican data are located in Fiz. 4.

S e- Locotion of Analyzed Deposits

Mexican Block

i ••,,,/../'Maya Block Chortis Block

Fig. 4. Distribution of deposits analysed in this study, as well as most of those analysed by Cumming et al. [37] from Mexico. The heavy dashed lines (1, 2, 3) separate crustal segments thought to have had separate plate tectonic histories in Mesozoic and Cenozoic time [30]. Areas A and C in Mexico contain deposits with lead isotopic compositions distinct from area B as noted in the text. Area E in Central America (the Chortis block) is isotopically similar to area B and could be a fragment of that terrain. 208 ing rocks, then perhaps similar patterns for oc- topic study of igneous rocks and ores from the Gold Hill eanic rocks yield similar information about ages of mining district, Utah, Utah Geol. 5, I (1979) I. source rocks, and do not necessarily result from 6 S.E. Kesler, Metallogenesis of the Caribbean region, .1. Geol. Soc. London 135 (1978) 429. simple mixing of two end members without time 7 J.R. Ladd, Relative motion of South America with respect significance. This interpretation would be con- to North America and Caribbean tectonics, Geol Soc. Am: sistent with Chase's [39] view that oceanic basalts Bull. 87 (1976) 869. can best explained in terms of two-stage models. 8 P. Molnar and L.R. Sykes, Tectonics of the Caribbean and In addition to providing insights into the source Middle America regions from local mechanisms and seismicity, Geol. Soc. Am. Bull. 80 (1969) 1639. of lead in the mineralization, our observations 9 H.H. Wilson, Cretaceous sedimentation and orogeny in appear to be useful to characterize crustal frag- Central America, Am. Assoc. Pet. Geol. Bull. 58 (1974) ments in the Mexican-Central American region. 1348. For instance, geologic and paleomagnetic data 10 W.A. Gose and D.K. Swartz, Paleomagnetic results from suggest that the region can be divided into at least Cretaceous sediments in Honduras: tectonic implications, Geology 5 (1977) 505 (see also discussions in Geology 6 4 crustal fragments, including the Maya and Chor- (1978) 440). tis blocks [30]. Considerable uncertainty persists 11 W.D. MacDonald, The Honduras paleomagnetic clock and over the provenance of the Maya and Chortis the Tertiary evolution of the Caribbean, EOS, Trans. Am. blocks, with the most commonly suggested possi- Geophys. Union 58, 6 (1977l 376. bility being detachment from western Mexico and 12 G. Dengo, Problems of tectonic relations between Central America and the Caribbean, G.C.A.G.S. Trans. 19 (1969) transport southward [40,41]. The similarity of lead 311. isotopic data from the Chortis and the Mexican 13 G. Dengo, Tectonic igneous sequence in Costa Rica, Bud- blocks, lends credence to this possibility. dington Vol., Geol. Soc. Am. (1962) 133. 14 D. del Guidice and G. Recchi, Geologia del area del proyecto minero de Azuero, Proyecto Minero Panama, Fas. I, Naciones Unidas, Panama (1969) 48 p. Acknowledgements 15 P.H. Mattson, Middle Cretaceous nappe structures in Puerto Rican ophiolites and their relation to the tectonic history of Laboratory work for this report was supported the Greater Antilles, Geol. Soc. Am. Bull 84 (1973) 1. by NSERC, and field work was supported by 16 P.H. Matts0n, Cuba, in: Mesozoic and Cenozoic Orogenic several private mining companies. Samples were Belts, A.M. Spencer, ed., Spec. Publ. Geol. Soc. London 4 (1973) 625. also contributed by D.J.T. Carson and R.H. Pem- 17 D.F. Cox, R.F. Marvin, J.W. M'Gonigle, D.H. McIntyre berton (Cuba), Charles L. Dahl (Haiti), George L. and C.L. Rogers, Potassium argon geochronology of some Riley (Costa Rica), M. Seaward (Honduras and metamorphic, igneous and hydrothermal events in Puerto Nicaragua) and D.M. Wyslousil (Guatemala). Rico and Virgin Islands, U.S. Geol. Surv. J. Res. 5 (1978) 689. 18 S.E. Kesler, J.F. Sutter, L.M. Jones and R.L. Walker, Early Cretaceous basement rocks in Hispanola, Geology 5 (1977) References 245. 19 K.M. Khudoley and A.A. Meyerhoff, Paleogeography and 1 G.L. Cumming and S.E. Kesler, Source of lead in Central geological history of Greater Antilles, Geol. Soc. Am. Mem. American and Caribbean mineralization, Earth Planet. Sci. 129 (1971) 199 pp. Lett. 31 (1976) 262. 20 J.E. Case, Crustal setting of mafic and ultramafic rocks and 2 B.R. Doe and R.E. Zartman, Plumbotectonics, I. The associated ore deposits of the Caribbean region, U.S. Geol. Phanerozoic, in: Geochemistry of Hydrothermal Ore De- Surv. Open-file Rep. 80-3041 (1980) 95 pp. posits, H.L. Barnes, ed. (John Wiley and Sons, New York, 21 S.E. Kesler and R. Ascarrunz-K, Lead-zinc mineralization N.Y., 1979) 2nd ed. in carbonate rocks, central Guatemala, Econ. Geol. 68 3 J.S. Stacey, R.E. Zartman and I.T. Nkomo, A lead isotope (1973) 1263. study of galenas and selected feldspars from mining dis- 22 S.E. Kesler, J.F. Sutter and M.J. Issigonis, Evolution of tricts in Utah, Econ. Geol. 63 (1968) 796. porphyry copper mineralization in an oceanic island arc, 4 R.E. Zartman, Lead isotope provinces in the Cordillera of Panama, Econ. Geol. 72 (1977) 1142. the Western United States and their geologic significance, 23 R. Weyl, Geology of Central America (Gebruder Econ. Geol. 69 (1974) 792. Borntraeger, Berlin, 1970) 2nd ed., 371 pp. 5 J.S. Stacey and R.E. Zartman, A lead and strontium iso- 24 H.L. Berryhill, Geology of the Ciales quadrangle, Puerto Rico, U.S. Geol. Surv. Bull. 1184 (1965) 116 pp. 209

25 R. Hey, Tectonic evolution of the Cocos-Nazca spreading 33 D.N. Gomberg, P.D. Banks and A.R. McBirney, Guatemala: Center, Geol. Soc. Am. Bull. 88 (1977) 1404. preliminary zircon ages from the central cordillera, Science 26 S.E. Church, The Cascade Mountains revisited: a reevalua- 162 (1968) 121. tion in light of new lead isotopic data, Earth Planet. Sci. 34 G. Salas, Carta y provincias metalogeneticas de la Repub- Lett. 29 (1976) 175. lica Mexicana, Cons. Rec. Miner. Publ. 21E (1975) 242 pp. 27 M. Tatsumoto, Isotopic composition of lead in oceanic 35 C.M. Tschanz, R.F. Marvin, B.M. Cruz, H.H. Mehnert and basalt and its implication to mantle evolution, Earth Planet. G.T. Cebula, Geologic evolution of the Sierra Nevada de Sci. Lett. 38 (1978) 63. Santa Marta, northeastern Columbia, Geol. Soc. Am. Bull. 28 G.L. Cumming, Lead isotope ratios in Deep Sea Drilling 85 (1974) 273. Project Leg 37 basalts, Earth Planet. Sci. Lett. 31 (1976) 36 C. Martin-Bellizzia, Edades isotopicas de rocas Vene- 1979. zolanas, Venzuela Bol. Geol. 10, 19 (1968) 356. 29 M. Yamaguchi, R.D. Russell and W.F. Slawson, Concentra- 37 G.L Cumming, S.E. Kesler and D. Krstic, Isotopic com- tion and composition of lead in Deep Sea Drilling Project position of lead in Mexican mineral deposits, Econ, Geol. Leg 37 cores, Can. J. Earth Sci. 14 (1977) 785. 74 (1979) 1395. 30 W.A. Gose, G.R. Scott and D.K. Swartz, The aggregation 38 T.E. Ewing, Lead isotope data from mineral deposits of of Mesoamerica: paleomagnetic evidence, in: The Origin of southern New Mexico: a reinterpretation, Econ. Geol. 74 the Gulf of Mexico and the Early Opening of the Central (1979) 678. North Atlantic Ocean, R.H. Pilger, ed., Louisiana State 39 C.G. Chase, Oceanic island Pb: two-stage histories and Univ., Dep. Geol., Publ. (1980) 51. mantle evolution, Earth Planet. Sci. Lett. 52 (1981) 277. 31 G.L. Cumming, J.S. Rollet, E.J.C. Rossotti and R.F. 40 D.E. Karig, R.K. Cardwell, G.F. Moore and D.G. Moore, Whewell, Statistical methods for the computation of stabil- Late Cenozoic subduction and continental margin trun- ity constants, I. Straight-line fitting of points with cor- cation along the northern Middle America Trench, Geol. related errors, J. Chem. Soc. Dalton Trans. (1972) 2652. Soc. Am. Bull. 89 (1978) 265. 32 R.H. Steiger and E. Jfiger, Subcommission on Geochronol- 41 B.T. Malfait and M.G. Dinkelman, Circum-Caribbean ogy: convention on the use of decay constants in geochro- tectonic and igneous activity and the evolution of the nology and cosmochronology, in: Contributions to the Geo- Caribbean plate, Geol. Soc. Am. Bull. 83 (1972) 251. logic Time Scale, Am. Assoc. Pet. Geol. Stud. Geol. 6 42 J.S. Stacey and J.D. Kramers, Approximation of a ter- (1978) 67. restrial lead isotope evolution by a two-stage model, Earth Planet. Sci. Lett. 26 (1975) 207.